Nouveau Recherche PDF, HTML, DOCX et bien d'autres formats.

Prépublication · 2026

Ground, Iterate, Trace: Decomposing AI Research Ideation

Samson Yu, Qiran Zou et al. — Monde

Source : arXiv

Résumé

Large language models have been widely used to generate plausible research ideas, yet it remains unclear what makes automated ideation effective and how generated novel proposals arise from prior scientific work. We introduce the Creative AI Researcher (CAR), an agentic framework combining literature grounding, iterative development and explicit conceptual operators. CAR constructs a semantic knowledge graph from retrieved literature and equips an ideation agent with graph-retrieval tools and four operators: bisociation, analogy, exploration and elaboration. Each operator call is checked again…

Actions

Citation